Chemoenzymatic synthesis is a preparative strategy which employs both chemical and biocatalytic steps in a reaction sequence. The biocatalytic transformations convert one organic compound to another by the use of enzymes, either isolated or as part of biological systems. These biocatalysts (enzymes) are in principle the same as any other type of catalyst. However, there are circumstances where these biocatalysts are especially useful, such as the induction of chirality due to enzyme enantiospecificity. These enzymatic reactions occur under mild conditions and are often more environmentally acceptable than classical chemical processes.
Lipases are the closest to optimum biocatalysts. They are isolated extracellular enzymes whose natural function is to hydrolyze glycerol esters. Many have wide substrate acceptability for ester hydrolysis, or, under the correct conditions, alcohol esterification. They are readily (and often cheaply) available and are experimentally simple, requiring no added cofactors and affording no side products. Not surprisingly these enzymes have been the most thoroughly studied for biocatalytic use in organic chemistry.
There are two types of substrate classes for lipase-catalyzed reactions. Meso or prochiral substrates constitute the first and most widely-studied class. The inherent chirality of the lipase distinguishes between two prochiral functions (esters or alcohols) on the same molecule to afford 100% conversion to (optimally) a single enantiomer.
The second class of substrates are the racemic systems, in which (optimally) only one of two enantiomers is recognized and hydrolyzed (or esterified) by the lipase, affording a 50% conversion to product and 50% recovered starting material of opposite configurations. This mixture must be physically separated to complete the enantiomeric differentiation. For substrates in which the acid rather than the alcohol portion is of interest, the separation is often possible by simple aqueous base extraction.
Alcohol-based substrates pose the most challenging separation problems due to the gross physical similarity between the alcohol and ester. It is to separations of this type that the present invention is directed.
Chemoenzymatic synthesis of optically active epoxybutadiene (hereinafter EpB) is a potentially attractive preparative method since a readily available source of EpB has recently become available. Novel, simple, and efficient preparations of optically pure C4 synthons derived from EpB would be synthetically useful, since most currently available chiral synthons have a three- or five-carbon backbone due to availability from natural sources. In fact, chain elongation of C3 synthons from the chiral pool currently comprises the major method for the preparation of optically active EpB and the corresponding diol (1,2-dihydroxy-3-butene).
For example, an early route to S-1,2-dihydroxy-3-butene and S-EpB relied on C6 D-mannitol (two identical three-carbon pieces) as the chiral starting material. (Baer, E.; Fischer, H. O. L. J. Biol. Chem. 1939, 128, 463) After formation of the terminal (symmetrical) diacetonide, the vicinal diol was oxidatively cleaved with lead tetraacetate to provide two molecules of the unstable acetonide of the three-carbon synthon R-glyceraldehyde. Wittig reaction with methylene triphenylphosphorane afforded 1,2-dihydroxybutene acetonide which was readily deprotected to the optically active 1,2-dihydroxybutene. Monotosylation of the diol and base treatment afforded optically active EpB. (Crawford, R. J.; Lutener, S. B.; Cockcroft, R. D. Can. J. Chem. 1976, 54, 3364.)
The corresponding R enantiomers were available from the antipodal three carbon synthon S-glyceraldehyde acetonide which has been prepared from L-ascorbic acid by several routes. After initial differential protection of the hydroxyl groups by sequential actonide formation and methylation, ozonolysis and lithium aluminum hydride treatment afforded S,S-1,2,3,4-tetrahydroxybutane 1,2-acetonide. Lead tetraacetate oxidative cleavage resulted in the desired S-glyceraldehyde acetonide. This material can be transformed to optically active 1,2-dihydroxy-3-butene and ultimately to R-EpB.
Alternatively, optically active 1,2-dihydroxy-3-butene can be prepared from one of the few four carbon synthons available from the chiral pool, tartaric acid. After preparation of the acetonide and reduction of the carboxyl groups, formic acid-induced rearrangement and hydrolysis of the resulting formates afforded the desired diol. This can be transformed to optically active EpB.
All routes suffer from synthetic problems. The oxidation steps mentioned above can be troublesome and produce highly toxic (lead) by-products. The first two routes also involve a cumbersome Wittig olefination of glyceraldehyde acetonide, itself a rather unstable species. In addition, each of the two routes can only be utilized for a single (but complementary) enantiomer due to the commercial availability of only D-mannitol and L-ascorbic acid. The route from tartaric acid is complicated by the formation of 1,4-dihydroxy-2-butene during the rearrangement reaction. Separation of this isomer from the desired 1,2-dihydroxy-3-butene is not trivial.
In actuality, only the route from tartaric acid is directed towards C4 synthons. The other schemes afford C4 materials as an afterthought by chain extension. A more direct approach, the synthesis of optically active C4 synthons from corresponding racemic C4 starting materials, would afford greater versatility for the preparation of diverse organic molecules. Therefore, the preparation of optically active EpB and derivatives (from racemic EpB) using biocatalysis technology is of great interest. An enantioselective lipase-catalyzedhydrolytic approach to this problem seemed promising due to the presence of diverse oxygen functionalities in many EpB derivatives.
EpB can be converted to a racemic ester by a number of routes. This ester is then subjected to enzymatic enantioselective hydrolysis to produce a mixture of enantiomerically enriched alcohol and enantiomerically enriched ester. While these compounds can be separated using chromatographic separation techniques, this is not practical on a large scale. Unfortunately, as mentioned previously, the separation of the alcohol from the ester is difficult because of the similarity of the physical characteristics of these compounds.
Thus, the present invention is directed to the problem of separating an optically active alcohol from a related optically active ester.